While all these experiments allow us to correlate different gene activities with the
induction of apoptosis, their exact contribution to apoptosis induction has to be
elucidated in every case. Starting with a similar approach to hunt for differentially
expressed genes, Michael Green and colleagues were able to attribute to the
upregulation of a lipocalin the cell death of a mouse pro-B cell line induced by
interleukin (IL)-3 deprivation (Devireddy et al., 2001). Lipocalins are a group of
small, mostly secreted proteins with various functions, ranging from the transport of
lipophilic molecules, prostaglandin synthesis, and the regulation of cell homoeostasis
to the modulation of the immune response (Flower, 1996). Transcending the mere
correlation of its upregulation, these workers demonstrated that this lipocalin is
necessary for apoptosis induction in their system, since antisense oligonucleotides
abrogated the effect. Furthermore, they showed that this molecule induces apoptosis
by an autocrine loop that is established in B cells by IL-3 deprivation. How the signal
of the proapoptotic lipocalin is transmitted in the cell is not yet known, although a
connection to the dephosphorylation of Bad, a Bcl-2-like protein, and its subsequent
activation was established.
Another interesting gene isolated with the correlative strategy is par-4, which was
found to be upregulated in apoptotic prostate cells (Sells et al., 1994). Par-4 features
a death domain and a leucin zipper, which is known as a protein-protein interaction
domain (Landschulz et al., 1988). Subsequent investigations showed that par-4 is
strongly expressed in a number of neurodegenerative diseases, such as Alzheimer’s,
Parkinson’s, and Huntington’s diseases, that are characterized by an excess of
apoptosis induction. Par-4 could mediate apoptosis by its interaction with a number
of apoptosis-regulating proteins, among them Bcl-2, protein kinase C zeta and
caspase-8 (Mattson et al., 1999). It has also been suggested that par-4 functionally
associates with NF-κB, an inducible transcription factor that can both repress and
mediate proapoptotic stimuli (Baichwal and Baeuerle, 1997).
Other investigators have used apoptosis induced by specific transcription factors
for the analysis with DNA microarrays or SAGE (serial analysis of gene expression).
Chiefly, p53, one of the most important tumor suppressor genes that can lead to
apoptosis, was studied (Kannan et al., 2001). Vogelstein and colleagues were able to
derive a general concept of p53-induced apoptosis from an overview of p53-target
genes (Polyak et al., 1997). They found a striking pattern of genes that can cause
oxidative stress, among them a galectin family member and a NADPH
oxidoreductase-related gene. In addition, they could repress apoptosis with
pyrrolidine dithiocarbamate (PDTC) and diphenyleneiodonium chloride (DPI), two
antioxidants. Consequently, these workers postulated that reactive oxygen species are
the determining factor in p53-activated cell death (Polyak et al., 1997). Other studies
by Arnold Levine’s group, also aiming at the transcription targets of p53, isolated
some, but not all, genes of the Vogelstein group (Zhao et al., 2000). These scientists
also determined cell-specific effects, the temporal succession of gene activation, and
various p53 stimuli. Over 100 genes were found to be up—or downregulated. They
belong to different classes such as growth factors, extracellular matrix proteins, and
adhesion molecules, but they also included genes, such as Fas and DR5, known to
202 GENETICS OF APOPTOSIS